© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 22-GEMMASCOPE-TRANSCODER-16K
    7. 12761
    Prev
    Next
    INDEX
    Explanations

    a wide variety of words that appear to be commonly used in academic/journalistic writing

    oai_token-act-pair · gemini-2.0-flash

    Abstract qualities

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     conseguenza
    -0.49
    biendo
    -0.41
    seguro
    -0.40
     législation
    -0.40
     goû
    -0.39
     ocasión
    -0.39
    ciclopedia
    -0.38
    Seguro
    -0.38
     véritables
    -0.38
     défaite
    -0.38
    POSITIVE LOGITS
     accuracy
    1.53
     simplicity
    1.49
     reasonableness
    1.48
     authenticity
    1.45
     robustness
    1.45
     flexibility
    1.45
     smoothness
    1.45
     reliability
    1.43
     transparency
    1.41
     stability
    1.41
    Activations Density 17.713%

    No Known Activations